Adherence to screening mammography recommendations in a university general medicine clinic

OBJECTIVE: To determine factors predicting adherence to a health care provider's screening mammography recommendation in a general internal medicine practice. DESIGN: Prospective observational study. SETTING: An urban academic general internal medicine practice. PATIENTS: Three hundred forty-nine asymptomatic women, aged 50 years and older, without prior history of breast cancer, who received a health care provider's recommendation for screening mammography. MEASUREMENT: Independent variables were: patient age, race, insurance type, educational level, and duration of affiliation with the practice; visit type; and health care provider gender and level of training. Dependent variables were acceptance of the recommendation and adherence, defined as undergoing mammography within three months of the recommendation. RESULTS: Overall, 193 (55%) of the women underwent the recommended mammography. Two hundred ninety-eight (85%) initially agreed to the recommendation, and of these, 190 (64%) completed mammography within three months. By univariate analysis, acceptance of the recommendation decreased significantly with increasing age (p<0.01), and by race (African-Americans 89% vs whites 82%, p=0.05). Only age remained independently predictive of acceptance in a multiple variable analysis. Among women who accepted the recommendation, adherence varied significantly according to race (white 70% vs nonwhite 59%, p=0.05), insurance type [Medicare as only insurance 45%, Medicaid 66%, non-health maintenance organization (non-HMO) private 62%, HMO 73%, p=0.03], and health care provider training (attending physicians 73%, residents 58%, nurse practitioners 47%, p=0.02). In a logistic regression analysis, insurance type and health care provider training remained independently predictive of adherence. CONCLUSION: Acceptance of screening mammography recommendations decreases with age. Among the women who agreed to the recommendation for screening mammography, insurance type and health care provider level of training best predicted adherence.